智能电网实现自愈控制的全局感知及风险驱动策略的理论研究

Self-healing, in essence, is well identified as the most desirable future of the smart grid, and it is the inheritance and development of the traditional preventive control, emergency control and recovery control in power system. The proposed research project which combined with the great strategic needs in the smart grid in the country, addresses global-senseing and risk-driven strategy in self-healing control methodology of the smart grid. The main content consists of three parts: Firstly, different control center in power system may have a different viewpoint in the face of the same physical phenomenons, since the physical phenomenon is global and the grid service data they get which may inaccuracy or deficiency is local, we need to utilize the technology, e.g. data mining, data fusion, study the consistency deceting model of the different control center, and put forward a global perception methodology of the smart grid.Secondly, all the blackouts have common traits, we need to explore the behavior of the cascading failure evolution according to the procedure of the the blackouts, build an online cascading failure risk assessment methodology, disclosure the relationship between the vulnerability and alteration in cascading failure, and then we can utilize the risk to identify the whole procedure of the blackouts. Thirdly, the concept of self-healing come from the biological field and the complex system, it is essentially the immune system, and it has different prototypes in the biological field and the complex systems. We can employ their ideas to manage the problems of risk supervision, risk constraint models and specification of the risk-fuzzy zones, furthemore investigate the conversion models in the hierarchy of smart grid among the self-adaptive preventive control, the self-adaptive emergency control and the self-adaptive recovery control, finally a more flatter risk-driven method in self-healing control theory of smart grid is presented. The proposed research project which strongly supports the large scale power system to achieve the goal of self-healing control functions, has important theoretical and practical significance.

自愈是智能电网最基本的特征，是传统电网安全运行预防控制、紧急控制和恢复控制的完善与发展。本项目密切结合我国大力发展智能电网的战略性决策，系统地研究智能电网实现自愈控制的全局状态感知及风险驱动策略的理论方法。内容包括：针对不同调度中心面对同一物理现象，可能作出不同判断的现象，建立调度中心的数据一致性检测模型，生成智能电网运行状态的全局感知方法；按照大停电事故演化过程，研究大电网连锁故障在线演化机理模型，提出在线生成在线评价的连锁故障实时风险评估方法，构建以风险为主线的大电网安全评价机制；借鉴生物和复杂系统等领域自愈的概念，探索自愈控制体系下的风险管理与约束模型及风险模糊地带的区分模型，研究自适应预防控制、自适应紧急控制与自适应恢复控制在风险驱动下的模态转化机制，提出更为扁平化的智能电网连锁故障自愈控制风险驱动策略。该项目将为大型互联电网实现自愈控制提供强有力的支持，具有重要的理论与现实意义。

自愈是智能电网的最基本的特征，是对传统电网的完善与发展。本项目针对智能电网研究了设备感知、风险感知以及风险评估一般性评估方法，包括含有特高压、新能源等特征的电网自愈特征进行了分析，研究了电网在故障过程中的风险的变化。主要的研究内容包括针对电网大停电过程中的连锁故障发展风险变化的特征进行分析，研究智能电网考虑配网保护装置时的网络攻击风险评估，研究含有智能配网与大量微电网之间的内部关联关系，研究含有分布式新能源的微网在信息安全下的风险分析等多个内容。其中连锁故障发展风险变化中考虑了电网中的断路器与继电保护特性，研究了故障的多个传播路径，提出了适合在线评估的连锁故障风险评估方法；在研究考虑保护系统时，通过对互感器、保护原理进行分析，获得网络攻击对电网安全性造成的影响;在研究配网与微网之间的关系时，通过分析蒙特卡洛分析方法提出了大量的微网并网可以有效改善电网的韧性；在研究新能源微网信息安全时，则以分析含有光伏、储能、风能的IIT微电网为例，讨论了网络攻击时的电网风险变化情况；通过分析继电保护装置的相互配合，提出了新能源微网的保护方法。